1. health and fitness
2. disorders
3. mental disorder
4. a.d.d.

1
Between persistence and flexibility:
The Yin and Yang of action control
Bernhard Hommel
Leiden University
Cognitive Psychology Unit & Leiden Institute for Brain and Cognition
Leiden, The Netherlands
Submitted April 20, 2015
In: Volume 2 (2015) of Advances in Motivation Science
Keywords: action control; cognitive control; will; automaticity; genetics; ego;
inhibition; persistence; flexibility; creativity; self-other integration
Correspondence:
Bernhard Hommel
Leiden University
Department of Psychology
Cognitive Psychology Unit
Wassenaarseweg 52
2333 AK Leiden, The Netherlands
e-mail: [email protected]
2
Content
1. Introduction
2. The will
3. The ego
4. Control as balance
5. The control state model
6. The plasticity of control states
7. Short-term biases of cognitive control
8. The impact of learning, experience, and culture on cognitive control
9. Conclusions
10. References
11. Acknowledgments
3
Abstract
Traditional approaches to action control assume the existence of a more or less unitary
control system that struggles with, and serves to overcome action tendencies induced by
automatic processes. In this article, I point out that and why these approaches fail to capture
the complexity and dynamics of cognitive control. I describe an alternative approach that
assumes that control emerges from the interaction of at least two counteracting forces: one
system promoting persistence and the maintenance of action goals and another promoting
mental and behavioral flexibility. I describe how this interaction might be shaped by various
factors, including genetic predisposition, learning, personal experience, and the cultural
context, and suggest a simple functional model (the Control State Model) that explains how
this shaping process works. Then I provide an overview of studies from various fields
(including perception, attention, performance monitoring, conflict resolution, creativity,
meditation, religion, and interpersonal perception and behavior) that successfully tested
predictions from the Control State Model.
4
1. Introduction
Human action is often goal-oriented (to a degree that the existence of goals is often
considered to be the defining aspect of human action), often initiated in the absence of actionrelated external stimulation, often spontaneous and non-habitual, and can commonly be
adjusted rapidly if no longer functional or not meeting situational challenges. The human
capacity to achieve all of this is commonly ascribed to what we now call cognitive control or
executive functions. The general idea is that the processes captured by these labels constitute a
kind of second layer of information processing: while the first layer consists of basic processes
to translate stimulation or input (the terms preferred by behavioristic and informationprocessing approaches, respectively)—such as sensory registration, identification, attentional
selection, response decision, and motor execution—the second layer operates on the first by
rewiring the respective processes in such a way that behavior is optimized and intended goals
are reached (e.g., Verbruggen, McLaren & Chambers, 2014).
Even though the terms cognitive control and executive function are relatively new (with
Atkinson and Shiffrin, 1968, as one of the first proponents), the underlying concepts and the
investigation of the underlying mechanisms is much older. In the following, I will discuss two
predecessors and the related theoretical implications that have survived the terminological
transition and that still dominate the discussion of human action control. Concepts that are used
for such a long time and that have stimulated so much research are unlikely to be entirely
incorrect, but I do think that they have limitations that need to be overcome in future research.
I will thus present an alternative view of human action control, at least of some relevant aspects
of it, that is also not without predecessors but has the advantage of providing a more dynamic,
and thus probably more realistic, view of how humans keep their actions adaptive. I will begin
by discussing two more traditional and popular views of action control—one assuming a will
that is fighting against habits and another assuming an ego that mediates between societal
5
requirements and personal urges—and point out a number of limitations of these views. Then
I will synthesize the positive aspects of these two views and bring them together with more
recent ideas from cognitive psychology and the cognitive neurosciences, to construct a more
dynamic approach to human action control. Finally, I will sketch major determinants of the
operation characteristics of control functions and provide some empirical examples to illustrate
and justify my choices. However, I will not start without emphasizing that this chapter is a
(empirical and theoretical) work in progress, a developing theoretical approach rather than a
complete theory.
2. The will
Approaches to human action control are older than the first psychological laboratory,
which explains why they were commonly focusing on the conscious experience of control—
the major empirical tool of non-experimental researchers. The dominant view during the years
in which the first laboratories were established was ideomotor theory, which has an even longer
philosophical tradition (Stock & Stock, 2004). The approach focuses on the process of gaining
control over one’s body and its movements. It basically assumes that perceptual impressions
of actively achieved changes in one’s environment are automatically associated with whatever
motor pattern was used to achieve them (for a review, see: Hommel, 2009). Once these
associations are created, the agent can reactivate the codes underlying the perceptual
impressions at will and thereby automatically prime the associated motor patterns—the motor
system thus becomes a slave of the will, as it were.
As the first experimental investigations of human cognition were focusing on
perception, ideomotor theory and the will in general did not receive substantial empirical
attention until the early 1900s. Around that time, experimental methods were developed to
assess the time it takes to implement a particular action goal (e.g., Michotte & Prüm, 1911) and
6
the impact of implementing a particular goal on information processing and conscious
experience (Ach, 1910; 1935). Of particular interest, even the very first attempts to assess
human will were using a scenario that survives to this day: Ach (1910) argued that the best and
most objective way to measure the human will requires putting it in opposition to a
counteracting force, which according to Ach was habit. Following this rationale, he developed
methods to strengthen particular stimulus-response associations through repetition (e.g., by
having participants produce rhyming responses to sequences of nonsense syllables) and then
tested the degree of willpower by requiring the participant to produce a different response on
old versus new items (for an overview, see Hommel, 2000a). As expected, items that were
previously associated with a different action category were more difficult and took more time
to respond to than new or familiar, but previously unassociated items. In numerous studies,
Ach and colleagues demonstrated that the size of this interference effect varies systematically
with the strength of the previously induced habit and/or the strength of the exerted or
individually available willpower (for an overview, see Ach, 1935).
Note the similarity between this approach and the now popular rationale to assess the
existence and efficiency of cognitive-control functions by analyzing how much they are
hampered by stimuli that are (or have been) associated with goal-incongruent responses—such
as in Stroop, Simon, or flanker tasks (e.g., Hommel, 2000b). In fact, it is not only the
experimental approach that is still heavily used, but the theoretical framework has remained in
place as well. It comes in several flavors. The revival at the end of the 1960s along the lines of
Atkinson and Shiffrin (1968) was basically replacing the apparently outdated terms “will” and
“habit” by the more depersonalized terms “cognitive control” (or executive control, or
executive function) and “automatic processes”. As in the original framework, novel and
unfamiliar actions were supposed to rely on control processes to configure the cognitive system
while overlearned actions could rely on automatic processes and thus require little or no
7
control. The basic idea was that learning is the mechanism that creates automaticity over time
by building more or less direct associations between stimulus representations and response
representations (e.g., Kornblum, Hasbroucq & Osman, 1990; de Wit & Dickinson, 2009).
There is a long-standing controversy with regard to the question how automatic processes or
automaticity need to be characterized, and it is fair to say that no universal criterion has been
found so far. Some authors have advocated non-intentionality as the key criterion while others
have suggested the ballistic, non-suppressible nature of automatic processes, and even others
have favored the lack of (commonly not further defined) “cognitive” or “central” resources
needed to run automatic processes or the immunity to interference from other processes (e.g.,
Bargh, 1994; Moors & de Houwer, 2006).
The major problem underlying discussions and theoretical considerations of that sort is
that they are facing a classical Rylean category issue (Ryle, 1949). This problem is a bit more
transparent when one refers to the original concept of the “will”. Will is something that persons
have but not a characteristic of a nervous system—much like there is nothing in a particular
building that reflects its being part of a university, as in Ryle’s famous example. Obviously,
the willing person has a nervous system and (if we accept the functionalist stance underlying
the rationale of cognitive psychology and the cognitive neurosciences) the process of willing
must be reflected in the activities of that system. Accordingly, it certainly makes sense to try
to capture the functional and/or neural reflection of the process of willing, but it makes little
sense to characterize particular processes as carrying or not carrying the essence of will. In fact,
the exact same process may very well subserve “will” in one situation and conflict with it in
another. Translating the personalized concept of will into the depersonalized concept of control
may sound more objective and scientific (Goschke, 2003) but it does not address the basic
logical problem.
Consider, for instance, automaticity in the Stroop task (Stroop, 1935). In this task,
8
participants are to name the color of color words, and they are known to be slower to do that if
the to-be-named color is inconsistent with the meaning of the word. The standard explanation
from control/automaticity approaches is that this is because people are used to reading words
rather than naming their color, which has automatized the reading process that now interferes
with the intentional color-naming response (for a review, see MacLeod, 1991). Note that in this
example the reading process is not uncontrolled: it is actually triggered, and thus under the
control of the word and its meaning. In that sense, the reading process is not automatic. One
might object that it is inconsistent with the action goal, and should in that sense be considered
automatic. But that does not hold as well. It is known that the Stroop effect is very small (less
than 10% of its original size) or absent in task versions operating with keypresses rather than
overt naming responses (McClain, 1983) and that the few demonstrations of reliable effects are
likely to result from internal naming (Martin, 1978, Chmiel, 1984). This implies that the
reading response is actually contingent on, and therefore controlled by the goal to utter color
words, which undermines the assumption that the reading response is automatic. How about
functionality, one might object: couldn’t one say that it is not useful to activate the reading
response related to the word meaning, and that the fact that we nevertheless find evidence for
such activation demonstrates automaticity? Even that approach is not tenable, as studies on the
so-called Gratton effect (Gratton, Coles & Donchin, 1992) have revealed. What these studies
show is that the interference related to the nominally irrelevant word meaning is drastically
reduced (if not eliminated) after incompatible trials, that is, after trials in which the word
meaning suggested the wrong response. In contrast, interference increases after compatible
trials, that is, after trials in which the word meaning suggested the correct response (e.g., Egner,
2007). But this means that the tendency to react to the nominally irrelevant stimulus feature is
closely depending on its informational value and, hence, on its functionality for action control.
And as this functionality depends on the action goal, which again is reflected in control
9
processes, we must conclude that what seems like the automatic activation of word meaning is
actually the goal-contingent, adaptive use of stimulus information that has just (i.e., in the
previous trial) proven to support action control.
These considerations suggest that it makes little sense to contrast control and automatic
processes and consider them to compete for action control, as the particularly popular dualprocess models in the field are doing (for an overview and a defense of the control/automaticity
dichotomy, see Verbruggen et al., 2014). Moreover, there is increasing evidence that what is
considered to be control processes actually shows many characteristics of what is considered
to be automatic. For instance, one of the key paradigms to investigate operations of cognitive
control is using task switching (for a review, see Kiesel et al., 2010). Participants are asked to
respond to particular stimuli according to one set of rules in some trials and according to
another set of rules in other trials. The standard finding is that performance is better if the set
of rules (i.e., the task) remains the same than if the task changes, which is commonly taken to
reflect the extra demands required to implement a different task set—a control process.
Interestingly, however, there is ample evidence that the implementation of tasks sets can be
triggered by external stimuli (e.g., Waszak, Hommel & Allport, 2003), even if these stimuli are
masked to render them unconscious (Reuss, Kiesel, Kunde & Hommel, 2011). Hence, there
are good reasons to assume that in a particular trial, “control/led” and “automatic” processes
are both under stimulus control (i.e., automatic) and contingent on the current action goal (i.e.,
controlled by “the agent”). This is consistent with approaches suggesting that planned action
consists in self-automatization. As suggested by Exner (1879), Bargh (1989) and others,
planning and preparing for an action comprises the implementation of all sorts of goalcontingent stimulus-response links that delegate the triggering and eventual execution of the
action to the environment as far as possible—a kind of prepared reflex (Hommel, 2000b).
To summarize, there is considerable evidence that automatic processes are under
10
surprisingly tight cognitive control while cognitive control processes can be triggered
automatically and even in the absence of conscious awareness. In principle, this makes it
impossible to identify clear-cut criteria to categorize processes according to that dichotomy,
which undermines its usefulness. The popularity of the dichotomy is likely to emerge from its
intuitive plausibility. Take, for instance, the example of substance addiction. Many will know
from their own experience that breaking with one’s habits can be extremely difficult, such as
quitting to smoke. Subjectively, the conflict emerges between the urge to smoke, especially in
the presence of smoke-compatible stimuli or situations, and the insight that leading a healthy
life requires one to quit. It is easy to see that this experience can be nicely couched in terms of
a struggle between will (the intention to quit) and habit (the overlearned urge to smoke).
However, it is a far cry from this everyday example expressed in everyday language to a
theoretical contrast between what dual-pathway theorists would consider the control process
(suggesting to quit) and the automatic process (suggesting to smoke). After all, the smoking
response meets all sorts of criteria of what in other circumstances would count as a control
process: it is expected to result and actually results in pleasure, fits with previous experience,
has shown to be functional in many earlier situations, fits with preferred personal stereotypes
(like admired actors that also smoked), and so forth and so on. Hence, there are many reasons
to smoke and it seems far-fetched to reduce the tendency to consider them a mere reflex—in
fact, even for those who want to quit, smoking is a goal-directed, intentional action. It is just
that the intention is inconsistent with other intentions—a classical goal conflict.
Taken altogether, a closer look reveals that the intuitive opposition of control and
automaticity makes little sense and fails to provide much help in characterizing and
categorizing the processes underlying human action control. The main difference between what
is commonly considered control processes and what is commonly considered automatic
processes lies in the familiarity with the stimulus-response relationship and the recency of
11
acquisition of the stimulus driving the action: while the stimuli driving the apparently more
“automatic” processes have commonly been acquired much earlier, the stimuli driving
“control” processes (i.e., the “if” parts of the instructed “if-then” rules) have been acquired or
instructed more recently. While this difference may allow for a relative categorization of one
process or stimulus-response association as more “automatic” than the other, it remains unclear
what benefits such a relative categorization brings theoretically.
3. The ego
Another approach to human action control is Freudian in nature. According to Freud’s
(1923) personality model, human action control is haunted by the continuous conflict between
societal rules and expectations (the Super-Ego) on the one hand and of personal needs and
urges (the Id) on the other. To deal with this conflict, humans develop an ego that mediates
between the two conflicting parties and tries to identify suitable compromises and solutions. In
many cases, acceptable solutions will consist in the inhibition of personal needs and urges,
which means that inhibition processes play a dominant role in this approach.
More modern versions of the ego model, such as the attentional approach of Kahneman
(1973) and the ego-depletion model of Baumeister, Bratslavsky, Muraven, and Tice (1998),
have focused on the assumed capacity requirements of cognitive control. The basic assumption
is that control operations require cognitive resources that automatic operations do not. This
allows for a number of empirical predictions that have been tested with some success. For
instance, extensive and repeated reliance on control operations should tend to exhaust the
hypothetical resource, which should impair subsequent control operations, but not automatic
operations, until the resource is refilled. Along the same lines, allowing the resource to refill,
or supporting and speeding up the refill process, should eliminate the reduced efficiency of
control operations.
12
The most problematic aspect of all available versions of ego or capacity theories is the
notorious under-definition of the most central concept—the resource. In the context of attention
theory, this has clearly stood in the way of a broader impact of the approach (Navon, 1984),
even though it is still used, often in a rather metaphorical way in applied fields (e.g., Wickens
1984). The main problem is that the failure to characterize the resource invites circular
reasoning (if some process unexpectedly fails to suffer from resource depletion it can be easily
declared automatic), which strictly speaking renders the approaches untestable (Navon, 1984).
A related problem refers to the nature of the capacity limitation. Straightforward predictions
only hold to the degree that the individual capacity limitation is fixed, so that the refill will
always re-establish the original or optimal level. However, some approaches have considered
the possibility that the individual capacity might vary (Kahneman, 1973). This is
motivationally not implausible and would be consistent with the “difficulty law of motivation”
proposed by Hillgruber (1912; see Hommel, Fischer, Colzato, van den Wildenberg, & Cellini,
2012). According to this law, the difficulty of an action is the motive for investing more effort
and devoting more cognitive control to reach the task goal. If so, increasing task difficulty
should automatically (“drive-like”, as Hillgruber says) increase cognitive control without
conscious deliberation, an idea that has lived on in several disguises (e.g., Kahneman, 1973;
Kukla, 1972; for a review, see Brehm & Self, 1989) and been confirmed empirically (e.g.,
Hommel et al., 2012). In a certain sense, the law can be considered a predecessor of the control
approach of Botvinick, Braver, Barch, Carter, and Cohen (2001), who suggested that
registering processing problems lead to a stronger focus on task-relevant information—a
classical control operation. However, while the assumption that individual capacity limitations
can vary is theoretically reasonable and empirically plausible, it basically renders ego-depletion
and other capacity models untestable.
Another Freudian approach, or family of approaches, focuses more on inhibition.
13
According to Freudian theorizing, socially inappropriate behavior would be prevented by
having the ego suppress corresponding drive-induced tendencies. Along the same lines,
numerous approaches have argued that, given the evidence that many stimuli can trigger
unwanted behavioral tendencies in conflict tasks (such as the reading response in the Stroop
task), there must be an inhibitory system suppressing these tendencies (e.g., Ridderinkhof,
Forstmann, Wylie, Burle & van den Wildenberg, 2010). In fact, many dual-pathway models of
action control have suggested the existence of such an inhibitory system, and individual
differences in the efficiency of the assumed inhibitory system have been taken to account for
various behavioral problems including addiction (for overviews, see Wiers & Stacy, 2006).
Behavioral studies investigating the efficiency of inhibition processes have often used conflict
tasks—following the rationale introduced by Ach (1910), with the assumption that more
evidence of conflict-induced interference from “automatic” processes indicates less
efficiency—and tasks assessing how fast people can abort an already planned or ongoing action
(e.g., Ridderinkhof et al., 2010).
Unfortunately, inhibitory approaches suffer from no less than three theoretical problems
that are commonly overlooked. First, assuming an inhibitory system that is independent from
the system representing the action goal doubles the “loans of intelligence” that theory building
should seek to minimize. As described by McCauley (1987), explaining the intelligence of
behavior by the intelligence of some assumed subsystem does not provide but merely postpones
an explanation. But “intelligence loans ought to be repaid!”, as McCauley rightly emphasizes.
To clarify, consider again our Stroop example. To account for the observation that participants
report the correct color name most of the time, even though they are more used to reading the
word, requires the assumption that the actual task or goal is represented somewhere in the
system and that it regulates information processing in such a way that the perceived color is
somehow translated into the appropriate color-naming response. This representation need not
14
be explicit but it must somehow incorporate the relevant stimulus-response rules and impact
information processing so as to apply them. Let us now change perspective and consider the
hypothetical inhibitory system. It may be easy to imagine that such a system can register the
presence of multiple conflicting response tendencies, but how can it know which one to inhibit?
To know that it would need to have access to a representation of the stimulus-response rules,
the ability to apply them to identify the correct response, and then use that knowledge to inhibit
all other response tendencies. In other words, the inhibitory system needs to be as intelligent
and knowledgeable as the goal system, which doubles the hypothesized intelligence. But if
nature has equipped us with one system that already knows what the right response should be—
the goal system as I call it—why would it provide us with an exact copy of it for inhibition
purposes? Why would it not leave the work to the first system, which already has all the
necessary information? Hence, assuming a separate inhibitory system strongly violates
Occam’s razor principle.
Second, almost all approaches to human decision-making share two basic assumptions
(for a review, see Bogacz, 2007): that decision-making is competitive (i.e., multiple
representations of alternatives actively compete for selection) and that it is selective and/or
capacity-limited (so that often one or few of the competing alternatives can be selected at one
time). Biologically plausible models commonly reflect these basic principles by assuming that
increasing the evidence for one alternative directly reduces the probability for the other
alternatives to be selected. In winner-take-all models, this is commonly translated into mutually
inhibitory relationships between representations of alternatives. Accordingly, increasing the
activation of the representation of one alternative would lead to the reduction of the activation
of alternative representations up to a point where the activation of one representation is at
maximum and the activations of all other representations is at some minimum or at least
sufficiently low. Note that this can be described as an inhibitory process: every gain for the
15
eventual winner implies a relative suppression of all losers. And yet, systems of that sort do
not require any independent inhibitor—inhibition is the necessary consequence of combining
competition with capacity limitations, as in every neural system. If so, the assumption of an
independent inhibitory system is entirely redundant and, thus, unnecessary.
Third, inhibitory approaches commonly jump from the observation of conflict-induced
processing costs to assumptions about the efficiency of inhibition. The idea is that, once
activated, incorrect response tendencies need to be suppressed, which takes time. As it is the
inhibitory system that is responsible for the suppression, the more time it takes, the less efficient
the inhibitory system must be. However, following the considerations just discussed, another,
more parsimonious interpretation is possible, if not more likely. If we combine the assumption
of competitive selection and capacity limitations, the time it takes to activate a correct response
against competing alternatives should depend on two factors: the degree to which the action
goal (or the representations embodying it) “supports” (provides top-down support) for the
representation of the correct response and the degree to which competing responses were
activated. Both of these factors are relatively well understood. Neuroimaging studies have
provided evidence that the degree of top-down support can fluctuate from trial to trial but that
it is strengthened whenever selection becomes more difficult (Egner & Hirsch, 2005). The
registration of such difficulty has been shown to increase the focus on task-relevant
representations in working memory (van Veen & Carter, 2002), which again results in
increased facilitation of goal-consistent representations—but not in increased inhibition of
irrelevant information (Egner & Hirsch, 2005). This suggests that the consideration of decisionmaking as competitive and capacity limited is sufficient to account for strategic adjustments in
action control, but leaves no space or need for a dedicated inhibitory system.
Along the same lines, a recent genetic study of Miyake, Friedman, and colleagues has
provided evidence that inhibition is an emergent property of cognitive processing rather than a
16
separable factor. In earlier approaches, Miyake et al. (2000) have suggested the existence of
three major systems underlying cognitive control: one responsible for information update,
another for shifting between tasks, and a third for inhibition. While the first two systems were
also indicated in the later genetic study (Friedman et al., 2008), the third, inhibitory system was
found to be more of a general characteristic of action control rather than a separable factor
equivalent to updating and shifting. This fits with the idea that inhibition is a necessary byproduct of competitive, capacity-limited selection.
4. Control as balance
An obvious commonality of many of the control approaches discussed so far is that they
assume one unitary system to take care of control. More recently, a number of authors have
argued for a more complex approach that considers multiple components or subsystems of
control. One influential example is the just-mentioned approach of Friedman, Miyake and
colleagues. From a more functional point of view, Goschke and colleagues (Goschke, 2003;
Dreisbach & Goschke, 2004) have pointed out that cognitive control often faces control
dilemmas that are binary in nature, with the choice between maintenance (or persistence) and
flexibility being a crucial one. While a strong maintenance of goals helps concentrating on
relevant information and suppressing irrelevant information, it increases the probability that
this renders a cognitive system too inflexible and insensitive to alternative possibilities and
reasons to give up ineffective action plans. In turn, focusing on flexibility facilitates switching
between alternative possibilities and actions but increases the possibility of distraction and
dysfunctional crosstalk between cognitive representations. The challenge for cognitive control
would thus be to find the right balance between persistence and flexibility for a given task and
situation. Indeed, Goschke and colleagues have shown how instructions and task conditions
can systematically bias control characteristics towards persistence or flexibility, and that doing
17
so creates the expected trade-offs. For instance, Dreisbach and Goschke (2004) found that the
induction of positive affect increases flexibility at the cost of persistence and Müller et al.
(2008) showed that prospective monetary gains have comparable effects. Given the link
between positive affect, reward expectations, and dopamine (Cools, 2008), Goschke and
colleagues argue that dopamine plays a crucial role in defining the control policy of executive
functions and shifting control between the poles of persistence and flexibility in particular.
This possible link between dopamine and control policies was also strongly emphasized
by Cools and D’Esposito (Cools, 2008, 2012; Cools & D’Esposito, 2010). According to their
approach, it is important to distinguish between at least two dopaminergic pathways—a
mesofrontal pathway targeting the prefrontal cortex and a nigrostriatal pathway targeting the
striatum. Prefrontal cortex and striatum are assumed to play opposite roles in determining
control policies, with the prefrontal component propagating persistence and the striatal
component supporting flexibility. This is consistent with numerous findings suggesting a
dominant role of the prefrontal cortex in behavior and cognitive skills requiring the
maintenance of information, such as working memory, and a dominant role of the striatum in
behavior and cognitive skills that require cognitive flexibility, such as in tasks calling for a
frequent update of information. A particularly important implication of the approach relates to
individual differences. As indicated by many pharmacological studies, neurotransmitters like
dopamine do not have linear performance characteristics (i.e., more is not necessarily better),
but follow an inverted U-shaped function, with best performance or efficiency related to
medium levels (Cools & D’Esposito, 2011). This means that interventions that are affecting
the current dopamine level are likely to have different implications for individuals with low
versus high baseline levels, and there is indeed evidence that many dopamine-targeting
interventions tend to “overdose” people with high dopamine levels. For instance, while positive
mood is commonly assumed to improve creativity (for a review, see Baas, De Dreu & Nijstad,
18
2008), this effect seems to be limited to individuals with low striatal dopamine levels (Akbari
Chermahini & Hommel, 2012a).
With a stronger emphasis on dopaminergic-receptor families, Durstewitz and Seamans
(2008) have proposed a dual-state framework. These authors focus on prefrontal cortex and
point out that the dopamine targeting this area is mainly processed by receptors from two
families: D1- and D2-class receptors. They discuss empirical and modeling evidence that the
prefrontal cortex can assume two different control states: One is characterized by the
dominance of D1 receptors, which is accompanied by high performance in tasks requiring the
maintenance of information. The other is characterized by the dominance of D2 receptors,
which is accompanied by high performance in tasks requiring flexibility in the switching
between options. The idea is that control policies are realized by shifting between D1- and D2dominated states of the prefrontal cortex, which amounts to finding the right balance between
persistence and flexibility.
Taken altogether, these available approaches differ in emphasis and level of analysis,
but they provide a nicely converging picture. All approaches emphasize the existence of more
than one executive control function, which goes beyond will- or ego-based unifactorial models.
This does not necessarily render these more simple models incorrect, it just suggests that they
underestimate the complexity and dynamical properties of cognitive control. Moreover, apart
from the more conceptually oriented psychometric approach of Friedman, Miyake and
colleagues, all approaches emphasize the interactive, to a considerable degree even
oppositional nature of the interactions between the components of cognitive control. This
suggests that control emerges from the competition between at least two systems, one
propagating persistence and the other propagating flexibility. The particular balance or
imbalance between these two systems determines the control policy, that is, the performance
characteristic of the control system. The more neuroscientifically interested approaches
19
consider dopamine the crucial mediator that, in interaction with existing dopamine levels,
control the relative impact of the two counteracting systems. One of these systems seems to be,
or at least includes the prefrontal cortex, which is targeted by the mesofrontal dopaminergic
pathway. The other system seems to be, or includes, the striatum targeted by the nigrostriatal
dopaminergic pathway. Interestingly, dopaminergic receptors are not equally distributed over
these two systems: While the prefrontal cortex is dominated by D1-family receptors, the
striatum is dominated by D2 receptors (Beaulieu & Gainetdinov, 2011). This renders the
approaches of Cools and D’Esposito (2010) and of Durstewitz and Seamans (2008) highly
compatible, arguably even functionally equivalent, as D1-dominance would directly translate
into a stronger relative contribution of the prefrontal cortex and D2-dominance into a stronger
relative contribution of the striatum.
5. The Control State Model
In the following, I will provide a brief overview of work led by Lorenza Colzato and
myself that was guided by the idea that cognitive control may not be a unitary function but
emerges from the interaction of two counteracting forces, which renders our approach a part of
a broader family of control accounts that also includes the discussed approaches of Goschke
and colleagues, Cools and D’Esposito, and Durstewitz and Seamans. Hence, even though most
of our studies were targeting a coarser level of analysis and were thus less committed to
particular brain areas and receptor families, we consider our general approach consistent and
compatible with these other approaches and we were greatly inspired by them.
Given that we were interested in bringing together phenomena related to cognitive
control from various levels of complexity, it was necessary to generalize and simplify the idea
of control as balance, and to make it more concrete at the same time. To achieve that, we tried
to relate this idea to existing models of human decision-making and considered how changes
20
in balance—that is, changes in control style from more persistence to more flexibility, and vice
versa—might affect decision-making processes. According to Bogacz (2007), there is a general
modeling trend from earlier evidence-collection models, which assume that representations of
to-be-decided-upon alternatives collect internal and external stimulus information speaking in
their favor, to more competitive models that reach decisions more quickly and that have greater
biological plausibility. Figure 1A shows the minimal requirements for a competitive model:
For one, the two or more alternative representations are competitive, which means that
increases in activation of one reduce the activation of the other (until one of them reaches a
certain threshold)—which in the figure is indicated by the mutually inhibitory links between
the alternatives. Note that the resulting competition amounts to a capacity limitation along the
lines discussed above, and that it leads to inhibition of less supported alternatives without the
need of having a separate inhibiting system. As alternatives may receive some support from
congruent internal or external sources, the dynamics of the interaction will have properties that
random-walk and diffusion models (e.g., Ratcliff & Rouder, 1998) are trying to capture. The
second ingredient that biologically plausible competitive models need is some specific support
from sources reflecting the current goal, which corresponds to what is considered top-down
support in many modeling approaches (e.g., Desimone & Duncan, 1995). It is this input that
makes sure that the “walk” is eventually not “random” but that the eventual decision is
consistent with the current goal, at least in most of the cases.
*** Figure 1 ***
How might a bias towards more persistence or more flexibility be integrated in such a
minimal model? According to Hommel and Colzato (2010) and Hommel (2012) there are at
least two, not necessarily mutually exclusive possibilities. One consists in increasing impact of
the goal to increase persistence (as indicated in Figure 1B) and reduce it to achieve flexibility
(Figure 1D). An increase will make it more likely that the most goal-consistent alternative will
21
eventually be selected and it will make it easier to suppress less consistent or irrelevant
alternatives, which would make decision-making more efficient, more selective, and more
focused. In contrast, a reduced impact of the goal will increase the possibility that goalinconsistent alternatives will be selected and lead to a stronger impact of irrelevant alternatives,
but it facilitates the switch to other alternatives and the consideration of multiple alternatives
in quick succession. The other possibility would be to strengthen (Figure 1C) or weaken (Figure
1E) the competitiveness between alternatives to achieve persistence and flexibility,
respectively. The consequences would be more or less the same as for the modulation of goal
impact, only that the modulation of competitiveness might introduce more goal-unrelated
influences on the activation of alternatives. As these two possibilities have turned out to be
difficult to disentangle empirically so far, and as they may very well be concurrently applied,
I will not try to discriminate between them in what follows. Instead, I will assume, for the sake
of simplicity, that changes of control states affect both the impact of the goal and the
competitiveness between alternatives, as indicated in Figures 1F and 1G. In the following, I
will use this model, which I will refer to as the Control State Model (or CSM), to explain how
situational circumstances, experience, learning, and people’s cultural context can bias cognitive
control states, and the information-processing characteristics they control, in predictable ways
and even create chronic biases that outlive the conditions for which they were originally
created.
6. The plasticity of control states
Task-switching studies have revealed that establishing a control state to orchestrate a
new task takes quite some time, often in the neighborhood of half a second or a second (for a
review, see Kiesel et al., 2010). Even though the duration of establishing a new control state
has been considered to depend on the number of to-be-changed control parameters (Logan &
22
Gordon, 2001) and the degree to which the new task set had been inhibited in earlier trials
(Mayr & Keele, 2000), most approaches implicitly or explicitly assume that the remains of
previous control states are fully eliminated before the new task will be executed. However,
there are some indications that this scenario is not realistic. For one, the first systematic studies
on task switching have already shown that previous control states can be sticky and affect later
control even after a new state has been implemented (Allport, Styles & Hsieh, 1994). For
another, there are some indications in the literature that parameters of previously established
states can systematically bias the characteristics of later states (Förster, 2012). These priminglike effects are particularly interesting from a CSM point of view, as I will elaborate in the next
section.
While short-term transfer of control-state characteristics from one task to another
supports the assumption that control states are inert, inertia by itself does not prevent random
and radical changes in control-state policies—it just slows them down. However, there is
considerable evidence that people do not choose control policies randomly or only according
to external demands; rather, they seem to have individual preferences for particular control
styles and default values that can be biased towards the persistence or the flexibility pole of the
control dimension that I have suggested. Some of these preferences are likely to reflect genetic
predispositions. There is considerable evidence that the genetic setup of individuals can have
specific, selective effects on the efficiency of processing along the mesofrontal or nigrostriatal
dopaminergic pathway, which has implications for people’s abilities to engage in persistence
or flexibility.
For instance, the COMT Val158Met polymorphism is known to affect frontal
dopaminergic processing, which again has been assumed to shift the dominance in
frontostriatal interactions towards the frontal system in some genotypes and towards the striatal
system in others (e.g., Nolan, Bilder, Lachman & Volavka, 2004). Indeed, individuals with a
23
genotype favoring frontal dopaminergic processing were shown to have more difficulties
switching from one task to another than individuals with a genotype favoring striatal
dopaminergic processing (Colzato, Waszak, Nieuwenhuis, Posthuma & Hommel, 2010).
Interestingly, the same polymorphism also predicted the degree to which people benefit from
playing videogames: individuals with a genotype favoring striatal dopaminergic processing
showed better transfer from a videogame training to task switching (Colzato, van den
Wildenberg & Hommel, 2014). Other polymorphisms are known to affect striatal dopaminergic
processing, which in turn is likely to shift the balance from or towards the striatal component
of the frontostriatal interaction from the other side. For instance, genetic markers of striatal
dopamine were shown to predict individual differences in the control of stimulus integration
over time (Colzato, Slagter, de Rover & Hommel, 2011), in the handling of integrated episodes
(Colzato, Zmigrod & Hommel, 2013), in attentional flexibility (Colzato, Pratt & Hommel,
2010) and functional impulsivity (Colzato, van den Wildenberg, van der Does & Hommel,
2010).
While it remains to be seen whether and to what degree these genetic constraints and
biases can be overcome, there is considerable evidence that personal experience, learning, and
cultural embedding can affect and create apparently rather stable inter-individual differences.
In the following, I will first focus on short-lived biases of control states that not only provide
evidence for the inertia of control parameters, but also demonstrate the usefulness of the CSM
in accounting for intra- and inter-individual differences in control policies under various
circumstances. Then, before I close with a brief conclusion, I will turn to more durable biases
and show that people can acquire particular personal control styles that predict important
characteristics of their performance.
24
7. Short-term biases of cognitive control
One set of studies that have tested predictions from CSM with respect to short-term
biases towards more maintaining or more flexible control states has used mood induction. The
induction of positive mood has often be assumed to promote “loose thinking” and creative
thought (Ashby, Isen & Turken, 1999), and there is considerable evidence that at least
brainstorming-like divergent-thinking tasks benefit from positive mood (Baas, De Dreu &
Nijstad, 2008; Isen, 1999). If so, one would expect that positive mood leads to a more “flexible”
control state as indicated in Figure 1D/E (cf., Nijstad, De Dreu, Rietzschel & Baas, 2010).
Take, for instance, the Alternate Uses Task (AUT) often used to assess divergent thinking
(Guilford, 1967). In this task, participants are presented with the description of a simple
everyday item, like a pen, and are then asked to list as many uses they can think of within
limited time. Such a task would require rather weak top-down support, as the search criterion
is rather vague. Given that many answers are possible and correct, the competitiveness between
alternatives should also be as weak as possible (i.e., just strong enough to select them
sequentially). This means that performance in this task would strongly benefit from a control
mode that is strongly biased towards flexibility. Accordingly, it makes sense that divergent
thinking benefits from positive-going mood.
These considerations are not limited to creativity tasks, so we can consider the
implications of positive mood for other tasks as well. If positive mood would bias control
towards more flexibility, it should increase the impact of goal-unrelated, irrelevant information
and reduce the impact of top-down control. This is consistent with observations of Dreisbach
and Goschke (2004), who found that the induction of positive mood is accompanied by less
systematic attentional focusing and greater distractibility. Along the same lines, there is
evidence that positive-mood induction reduces the degree to which task performance is
monitored and adjusted in the face of internal conflict. It is known that, in tasks with conflicting
25
stimuli, like Stroop or flanker tasks, the experience of conflict leads to a reduction of the impact
of irrelevant information in the next trial—the Gratton effect (Gratton, Coles & Donchin,
1992). According to Botvinick and colleagues (2001), this is because the registration of
multiple response activations (i.e., response conflict) is fed to frontal systems responsible for
the maintenance of the action goal, and results in the reactivation/strengthening of the goal
representation. This should increase the top-down control of information processing in the next
trial, and thereby reduce possible conflict. Interestingly, the presentation of unexpected reward
(van Steenbergen, Band & Hommel, 2009, 2012) and the induction of positive mood (van
Steenbergen, Band & Hommel, 2010) effectively prevent this process of control adjustment,
suggesting that positive mood indeed reduces the impact of top-down control. A recent
neuroimaging study indicates that this reduction is caused by attenuating the response of the
assumed conflict monitor (the anterior cingulate cortex) to stimulus-induced conflict (van
Steenbergen, Band, Hommel, Rombouts & Nieuwenhuis, in press). While positive mood
reduces top-down control, negative mood leads to an increase: Van Steenbergen, Booij, Band,
Hommel, and van der Does (2012) tested the effect of dysphoria induced by acute tryptophan
depletion in remitted depressed patients and observed a significant increase of control
adjustment under these conditions.
Systematic mood effects were also obtained in dual-task contexts. For instance, Zwosta,
Hommel, Goschke, and Fischer (2013) reported that the induction of negative mood led to a
stronger shielding of the prioritized task (i.e., less crosstalk) than the induction of positive
mood, which fits with the idea that positive mood reduces top-down impact. Fischer and
Hommel (2012) used a different rationale to investigate dual-task performance. They mixed
the task with a creativity task that required either divergent thinking, as in the mentioned AUT,
or convergent thinking, as in the Remote Associations Task (RAT; Mednick, 1968). This task
provides more and rather tight top-down constraints, and there is only one possible answer per
26
item (e.g., which word goes with “market”, “man”, and “glue”?), suggesting that the task calls
for a control state with a strong impact of the goal—a bias towards persistence. If so,
performing the RAT should increase top-down control and reduce crosstalk in an overlapping
dual-task, which is indeed what Fischer and Hommel observed.
So far, I have considered mood as an independent variable that biases cognitive control
to a more persistence-supporting or a more flexibility-supporting state and task-related controlstate configurations as the dependent variable. From a functional perspective, one might
consider influence in the opposite direction as well. That is, adopting a control state close to
the persistence pole of the control dimension might be associated with more negative mood
than adopting a control state close to the flexibility pole. Akbari Chermahini and Hommel
(2012) tested this possibility by having participants perform convergent- or divergent-thinking
tasks and assessing whether these tasks had an impact on mood. This was indeed the case:
mood was improved by performing the divergent-thinking task and performing the convergentthinking task had the opposite effect.
Another factor that has been found to have a systematic impact on cognitive control
states is meditation. Meditation has been notoriously suspected to improve creativity, but
systematic evidence regarding whether this is really the case is rare. We have pointed out that
this is likely to be the case for two reasons (Lippelt, Hommel & Colzato, 2014). For one,
different kinds of meditation have different goals and use different techniques to achieve them.
As pointed out by Lutz, Slagter, Dunne, and Davidson (2008), some techniques are coaching
the meditator how to focus on one single thought while avoiding distraction (focused-attention
meditation), while other techniques aim at teaching the meditator how to open up and to accept
any thought that may come up at any time (open-monitoring meditation). It is unlikely that
these two techniques are associated with the same cognitive-control state; rather, it makes sense
to assume that focused-attention meditation serves to bias cognitive control towards the
27
persistence pole while open-monitoring meditation promotes flexibility (Lippelt et al., 2014).
For another, creativity is not a unitary construct and comprises several components with
partially opposite characteristics (Wallace, 1926; Guilford, 1967). In particular, some creativity
tasks (like the RAT) are tapping more into convergent thinking while others (like the AUT) are
tapping more into divergent thinking. If we consider the above reasons and evidence that
convergent thinking benefits from a persistence-heavy control state while divergent thinking
benefits from a flexibility-biased state, it makes little sense to assume that both kinds of
thinking benefit from the same kind of meditation. Rather, focused-attention meditation should
promote convergent thinking while open-monitoring meditation would be suspected to
facilitate divergent thinking.
These predictions were indeed confirmed in studies that systematically compared the
impact of focused-attention meditation and open-monitoring meditation on convergent and
divergent thinking. In a study on meditation practitioners (Colzato, Ozturk & Hommel, 2012)
and a study comparing practitioners with meditation novices (Colzato, Szapora, Lippelt &
Hommel, in press), open-monitoring meditation led to a substantial improvement of divergent
thinking without affecting convergent thinking. Focused-attention meditation in turn tended to
improve convergent thinking without having any impact on divergent thinking. Why the impact
of focused-attention meditation on convergent thinking was not statistically reliable is likely to
due to the fact that most participants were students—who can be assumed to engage in
convergent thinking on an everyday basis -- so that their performance was probably at ceiling
already.
A further line of research investigated whether control states might affect interpersonal
processes. The rationale motivating this research is based on the assumption that people
cognitively represent themselves and others like any other event (Hommel, Müsseler,
Aschersleben & Prinz, 2001; Hommel, 2004): as a network of codes that represent various
28
perceptual and conceptual features characterizing the event (for further discussion, see
Greenwald et al., 2002; Hommel, 2013; Hommel, Colzato & van den Wildenberg, 2009; Kim
& Hommel, in press). If this is the case, one would assume that a control state close to the
persistence pole of the dimension should lead to a more pronounced discrimination between
“me” and “other”, because this control state is associated with maximal competitiveness
between alternative representations. In other words, the mutually inhibitory links between the
“me” representation and the “other” representation should be particularly strong, so that
activating one should tend to inhibit the other to a maximal degree. In contrast, a control state
close to the flexibility pole should lead to minimal discrimination, that is, to maximal
interpersonal integration. In other words, this control state would propagate a view of me and
other as one unit while a persistence-heavy control state would induce a more “individualistic”
representation.
Along these lines, one would expect that engaging in activities that are likely to drive
the control state towards persistence would reduce self-other integration, while engaging in
flexibility-heavy tasks should have the opposite effect. This pattern of results was indeed
observed by Colzato, van den Wildenberg, and Hommel (2013): Performing a divergent
thinking task was associated with a substantial increase in the Joint Simon effect (Sebanz,
Knoblich & Prinz, 2003; for an overview, see Dolk et al., 2014), which indicates the degree to
which participants are affected by the presence of another person working on the same task.
The same effect was obtained by having participants to circle either personal pronouns (such
as “I,” “my,” or “me”) or relational pronouns (such as “we,” “our,” or “us”) in a text (Colzato,
de Bruijn & Hommel, 2012). In another study, engaging in divergent thinking was reported to
increase interpersonal trust (Sellaro, Hommel, de Kwaadsteniet & Colzato, 2014), and the same
effect was obtained by exposing participants to the scent of lavender, which is assumed to
induce a more flexibility-heavy control state (Sellaro, van Dijk, Rossi Paccani, Hommel &
29
Colzato, 2014).
8. The impact of learning, experience, and culture on cognitive control
The evidence I have discussed so far shows that particular control states have a specific
impact on internal representations and information processing. This impact was restricted to a
temporal scale of seconds or minutes, and this is indeed a scale that is particularly likely to
show such effects. As it is known from task-switching studies (see above), changing control
states takes quite a bit of time and just-implemented states tend to be sticky to some degree, so
that it makes sense to look for state-related aftereffects and crosstalk in in the range of minutes
or seconds. And yet, it also makes sense to assume that longer-term effects and biases are
possible. Consider the hypothesized control dimension ranging from extreme persistence and
top-down impact to extreme flexibility and openness. There are certainly tasks and situations
in life that call for control states that are more biased towards one than the other pole, but it is
unlikely that there is a particular optimal value for each task or situation. As emphasized by
Goschke (2003), some degree of goal-orientation/top-down control and some degree of
flexibility will be necessary under almost all circumstances, which still leaves quite some range
of possible values on the hypothetical dimension—not unlike the case of choosing between
speed and accuracy. The existence of a range of possibilities creates uncertainty, which is
commonly assumed to be aversive. This suggests that people seek ways to reduce uncertainty
regarding control parameters. One way to reduce it is to adopt a particular default value, which
an individual can use in all situations that do not require deviations from that value. Where do
individual default values come from? Several options are possible and they are not mutually
exclusive. In the following, I will discuss a few options that colleagues and I have addressed
empirically.
One relates to personal conditions that are more compatible with one than with the other
30
pole of the hypothetical control dimension. Engaging in a loving relationship is such a
condition, as it is likely to promote positive feelings and general positive mood. If so, CSM
would predict that passionate lovers are less efficient in maintaining tight action control.
Indeed, van Steenbergen, Langeslag, Band, and Hommel (2014) observed that the intensity of
passionate love as assessed by the Passionate Love Scale predicted the individual efficiency in
cognitive control as measured in Stroop and flanker-task performance—with greater passion
leading to less control.
Another way to influence cognitive control is learning. Colzato, van Leeuwen, van den
Wildenberg, and Hommel (2010) considered that the new generation of “First Person Shooter”
games may require players to develop a more flexible mindset that allows them to engage in
complex scenarios, to rapidly react to moving visual and sudden acoustic events, and to switch
back and forth between different subtasks. If so, one would expect that extensive experience
with such games should enhance cognitive flexibility and, indeed, players of such videogames
were found to be superior to non-players in a task-switching paradigm. While this study does
not reveal whether and how performance in persistence-heavy tasks was affected, it does
suggest that the flexibility part of control can be systematically trained. However, it must be
noted that the study was correlational in nature, which leaves the possibility that players are a
self-selected group with superior flexibility skills. It is also important to point out that taskswitching paradigms are rather complex, and that task-switching performance arguably
comprises both persistence and flexibility components.
Another example of learning-related changes in cognitive-control abilities relates to
bilingualism. There is considerable evidence that bilinguals not only outperform monolinguals
in various language tasks (even though bilinguals can show poorer performance early in
learning), but they show superior performance in a number of non-lingual tasks as well (for an
overview, see Bialystok & Craik, 2010). These observations have motivated the view that
31
acquiring a new language goes hand-in-hand with learning how to suppress conflicting
information. In the learning process, this information will be mainly linguistic, such as the
words and grammatical structures of one’s native language. However, if we assume that
discriminating between, and holding in check conflicting languages is a general control
process, it is very well possible that training and improving this process generalizes to the
control of the processing of non-linguistic material.
In a systematic comparison of tasks tapping into direct, “active” suppression (like the
Stop-Signal task: Logan & Cowan, 1984) and tasks involving the inhibition of alternatives
through the process of competitive selection, as discussed above, Colzato et al. (2008) were
able to show that it is only the latter, not the former process that bilinguals excel in. For
instance, while bilinguals and monolinguals showed no difference whatsoever in the stopsignal task, bilinguals performed more poorly in the attentional blink task. In the attentional
blink task, participants report the presence and identity of two targets presented in close
succession. The term “attentional blink” refers to the common observation that both targets can
be accurately reported with longer lags between them while the second target is often missed
when the lag is short. This has been attributed to the processing of the first target, which seems
to be so demanding that a second target cannot be considered before processing the first is
completed (for an overview, see Hommel et al., 2006). Interestingly, however, there are
considerable interindividual differences in the size of the effect (e.g., Martens, Munneke, Smid
& Johnson, 2006) and situational circumstances under which the “blink” of the second target
can be avoided altogether (e.g., Olivers & Nieuwenhuis, 2006). Moreover, trials in which the
second target is missed are characterized by a particularly large allocation of attentional
resources to processing the first target (Shapiro, Schmitz, Martens, Hommel & Schnitzler,
2006), which suggests that the attentional-blink phenomenon is due to too much top-down
control (Olivers & Nieuwenhuis, 2006). If so, it makes sense that bilinguals, who have been
32
claimed to acquire stronger top-down control (e.g., Bialystok,Craik, Klein & Viswanathan,
2004), exhibit a more pronounced attentional blink. It also makes sense that bilinguals
outperform monolinguals in convergent thinking, but show poorer performance than
monolinguals in divergent thinking (Hommel, Colzato, Fischer & Christoffels, 2011).
Becoming and remaining a bilingual is certainly a long-term learning experience, but
the learning process and the exercise is commonly much more socially embedded than
mastering a videogame. Given that the learning experience seems to affect people’s personal
control styles, we can hypothesize that societal practice and culture might contribute to the
development and persistence of cognitive control. We pursued this possibility in a recent line
of research that looked into the impact of religion on cognitive control. There is some evidence
from cultural studies that control styles might be sensitive to the cultural environment in which
people grow up. For instance, people growing up in North America have been shown to be less
sensitive to contextual cues and show a more analytic cognitive style (i.e., to pay more attention
to local features of objects and events) than people growing up in Asia, who exhibit a more
holistic style (i.e., pay more attention to global features)(Nisbett & Miyamoto, 2005). The
concept of culture is very hard to define, however, so that one for instance can argue whether
all citizens of the United States really share the same culture and whether one can speak of a
homogeneous Asian (or even “Eastern”) culture. A similar concept that is easier to define is
religion, and here I do not mean religiousness as such, but adopting a particular religious faith.
Even though not all faiths are based on available holy scripts and rules, and even though the
degree to which members of a religious community follow religious rules and regulations
varies, some basic and widely shared rules and values can often be identified for at least some
variants of religious conviction.
How might religion shape cognitive control? Hommel and Colzato (2010) have argued
that many kinds of religious faith have characteristics that are likely to require a particular
33
cognitive-control style, so that adopting a particular faith is likely to lead to the establishment
of control-parameter defaults that are compatible with the required style. For instance, Dutch
neo-Calvinism is built upon the concept of sphere sovereignty, which emphasizes that each
sphere or sector of society has its own responsibilities and authorities, and stands equal to other
spheres (e.g., Boesak, 1984). This implies that active Dutch Calvinists have been often socially
rewarded for adopting a control style that emphasizes a rather independent view of the self.
According to my reasoning, this should require a pronounced bias towards the persistence pole
of the control dimension, which again implies strong reliance on top-down control, good
shielding against irrelevant information and little distractibility, but also comparably less selfother integration and global attention.
In a first test of these ideas, we (Colzato, van den Wildenberg & Hommel, 2008)
compared Dutch Calvinists with Dutch atheists that were brought up in the same country and
culture and were matched with respect to race, intelligence, sex, and age, in the global-local
task developed by Navon (1977). This task assesses how fast people can process global and
local characteristics of hierarchically constructed visual stimuli (e.g., larger symbols made of
smaller symbols). Typically, this task gives rise to the “global precedence” effect, which
demonstrates that global features can be processed faster than local features—presumably due
to a mixture of physiological and attentional factors. We were particularly interested in the size
of this precedence effect. If Calvinists would be more oriented towards local details, their
precedence effect should be small, meaning that they should not be much slower to report local
than global stimulus characteristics. In comparison, atheists should show a larger precedence
effect, that is, a more pronounced delay in reporting local as compared to global characteristics.
This is indeed what we observed. A follow-up study replicated the effect and showed that the
attentional control characteristics of baptized atheists (i.e., individuals raised as Calvinists, but
no longer believing or engaging in religious practices) are indistinguishable from that of
34
practicing Calvinists (Colzato, van Beest, van den Wildenberg, Scorolli, Dorchin, Meiran,
Borghi & Hommel, 2010). This suggests that early practice in living a particular faith is more
important than continued practice—chronic control parameters seem sticky. The same study
also demonstrated that Italian Roman Catholics and Israeli Orthodox Jews (i.e., members of
religions that emphasize social responsibility and interrelatedness of all societal aspects) show
the opposite effect than Calvinists do (i.e., the global precedence effect was more pronounced
in the religious groups than in matched atheist groups), as our approach would predict. Along
the same lines, Colzato, Hommel, van den Wildenberg, and Hsieh (2010) reported a more
pronounced global precedence effect in Taiwanese Buddhists (i.e., members of a religion that
also emphasizes social interrelatedness) than in matched Taiwanese atheists. Taken together,
these observations rule out that religiousness as such, or some correlated behavioral or
personality feature accounts for our findings. Instead, it seems to be the particular religious
practice that creates chronic control-parameter values (Hommel & Colzato, 2010).
A further study showed that religious faith not only affects attentional control, but the
sensitivity to task-irrelevant information as well. According to the CSM, Calvinists would be
expected to show particularly low sensitivity to task-irrelevant information while Catholics
should show particularly high sensitivity. Hommel, Colzato, Scorolli, Borghi, and van den
Wildenberg (2011) tested this prediction in a Simon task that assesses the degree to which
response selection is affected by irrelevant stimulus information. While Dutch Calvinists
showed a less pronounced Simon effect than matched Dutch atheists, Italian Catholics showed
a more pronounced Simon effect than Italian atheists—which confirms expectations. A further
study tested the prediction that Calvinists would show a particularly large attentional blink.
Recall that the attentional blink has been attributed to too much top-down control, suggesting
that religious practice that emphasizes top-down control, as Dutch Calvinism does, should
increase the effect. This is indeed what Colzato, Hommel, and Shapiro (2010) demonstrated.
35
Finally, Colzato, Zech, Hommel, Verdonschot, van den Wildenberg, and Hsieh (2012)
investigated the impact of Buddhism on self-other integration. As explained in the previous
section, self-other integration can be expected to increase to the degree that control states
approach the flexibility pole. If so, practicing Buddhists would be expected to show more
pronounced self-other integration than atheists, and one major aim of Buddhist practices is
indeed to overcome the separation of self and other (Dogen, 1976). Colzato et al. tested this
hypothesis by comparing Taiwanese Buddhists with matched Taiwanese atheists in a Joint
Simon task. As predicted, Buddhists showed a more pronounced effect, suggesting stronger
self-other integration.
As pointed out already, religious practice is an example of a relatively well-defined
social activity that is likely to shape the behavior of the involved individual. But it is not the
only example. Another is membership in a relatively coherent minority group. To investigate
whether such membership can also have a long-lasting impact on cognitive control, we
compared homosexual and heterosexual individuals in a global-local task (Colzato, van
Hooidonk, van den Wildenberg, Harinck & Hommel, 2010). Our theoretical vantage point was
speculation about the possible existence of some kind of “gaydar” (Reuter, 2002), an assumed
special skill of homosexual individuals to detect the sexual orientation of others. Previous
studies have revealed that homosexual participants are indeed better than heterosexuals in
detecting the sexual orientation of strangers shown on photographs or in short movie clips, but
this advantage disappears with slightly longer movie clips (Ambady, Hallahan & Conner,
1999). We took this outcome pattern to rule out differences in knowledge about visible cues
signaling sexual orientation, such as clothing, hair dress, or gesture. Instead, homosexuals
might check for such cues more actively, as they rely more on them to identify possible
partners. If so, the default control setting of homosexuals might be more biased towards local
aspects of stimulus events, as this is the setting they are using more often than heterosexuals—
36
at least for the purpose of partner identification. We tested this possibility by presenting
homosexuals and heterosexuals with the global-local task. As expected, homosexuals had a
smaller global preference effect than heterosexuals, which indicates that they have a relative
preference for the processing of local stimulus aspects.
9. Conclusions
The main aim of this article is to point out that action control is unlikely to be a unitary
function, but rather emerges from the interaction of two counteracting forces, one propagating
a strong top-down control of information processing and the other opening up the system for
alternative options and novel information. As I have argued, this more dynamic approach has
functional and neural plausibility, and is consistent with recent functional and neuroscientific
insights. In this article, I did not focus on the neuroscientific evidence, but rather tried to show
that a relatively simple functional model is sufficient to account for various phenomena
indicating the existence of short-term biases of cognitive-control styles and long-term chronic
biases associated with particular kinds of experience and cultural context.
Even though the CSM I proposed has two components, these components are different
from those considered by traditional dual-pathway models. The latter assume that the two
alternative or conflicting pathways differ with respect to their reliance on familiarity and
previous experience, their controllability, and their mechanisms. The CSM does not assume
any of those differences and it in particular does not consider a stronger contribution of the
non-frontal component reflecting less control. Nor does it consider control configurations that
leave space for other than goal-related top-down information necessarily inefficient or
dysfunctional. The idea that more top-down control is better, as implicit or even explicit in
many dual-pathway approaches, is in fact not particularly realistic outside of the psychological
laboratory. In the absence of an experimenter who has already limited the available possibilities
37
to exactly one task, one or two stimuli, and one or two predefined responses mapped upon
them, a human agent is always facing the possibility that the current task is not the most
important, most appropriate, most efficient, and most successful, which requires the continuous
monitoring for interesting alternatives. Switching off processing channels busy with these
considerations might buy a few milliseconds, but could have fatal consequences.
CSM can also do without an ego or agent, and without a dedicated inhibitory system.
Inhibition emerges through competition and, thus, does not require loans of intelligence to
explain how inhibitory systems know which representations to target. The main intelligence in
the control system that calls for explanation is the choice of control styles: why does control
sometimes lean towards persistence and sometimes towards flexibility, and what are the criteria
for changing this policy? I have tried to reduce this loan of intelligence as well by considering
genetic reasons for individual differences and by showing that individuals can acquire chronic
default values through cultural learning and other kinds of personal experience. The remaining
flexibility that people do seem to have might be explained by considering situational cues and
immediate feedback, but certainly more research will be necessary to unravel the details of
these processes. Moreover, it is fair to say that the CSM is currently more of a heuristic tool
rather than a formalized theoretical framework, and making progress on this will require much
more fine-tuning. Nevertheless, I am convinced that considering the dynamic, interactive
nature of cognitive control will deepen our insight into the mechanisms underlying the
emergence of human behavior.
10. References
Ach, N. (1910). Über den Willensakt und das Temperament [on act of will and
temperament]. Leipzig: Quelle & Meyer.
Ach, N. (1935). Analyse des Willens [analysis of the will]. In E. Abderhalden (Ed.),
38
Handbuch der biologischen Arbeitsmethoden, Vol. VI. Berlin: Urban & Schwarzenberg.
Akbari Chermahini, S., & Hommel, B. (2012a). More creative through positive mood?
Not everyone! Frontiers in Human Neuroscience, 6:319.
Akbari Chermahini, S., & Hommel, B. (2012b). Creative mood swings: Divergent and
convergent thinking affect mood in opposite ways. Psychological Research, 76, 634-640.
Akyürek, E.G., Riddell, P.M., Toffanin, P., & Hommel, B. (2007). Adaptive control of
event integration: Evidence from event-related potentials. Psychophysiology, 44, 383-391.
Akyürek, E.G., Toffanin, P., & Hommel, B. (2008). Adaptive control of event
integration. Journal of Experimental Psychology: Human Perception and Performance, 34,
569-577.
Allport, D.A., Styles, E.A., & Hsieh, S. (1994). Shifting intentional set: Exploring the
dynamic control of tasks. In C. Umilta & M. Moscovitch (Eds.), Attention and performance,
Vol. XV (pp. 421–452). Cambridge: MIT Press.
Ambady, N., Hallahan, M., & Conner, B. (1999). Accuracy of judgments of sexual
orientation from thin slices of behavior. Journal of Personality and Social Psychology, 77,
538–547.
Ashby, F.G., Isen, A.M., & Turken, A.U. (1999). A neuro-psychological theory of
positive affect and its influence on cognition. Psychological Review, 106, 529–550.
Atkinson, R.C., & Shiffrin, R.M. (1968). Human memory: A proposed system and its
control processes. In: K.W. Spence & J.T. Spence (eds.), The psychology of learning and
motivation, Vol. 2 (pp. 89-195). New York: Academic Press.
Baas, M., De Dreu, C.K.W., & Nijstad, B.A. (2008). A meta-analysis of 25 years of
research on mood and creativity: hedonic tone, activation, or regulatory focus? Psychological
Bulletin, 134, 779–806.
Bargh, J.A. (1989). Conditional automaticity: Varieties of automatic influence in social
39
perception and cognition. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 3–
51). New York: Guilford Press.
Bargh, J.A. (1994). The four horsemen of automaticity: Awareness, efficiency,
intention, and control in social cognition. In R.S. Wyer, Jr., & T.K. Srull (Eds.), Handbook of
social cognition (2nd ed., pp. 1-40). Hillsdale, NJ: Erlbaum.
Baumeister, R.F., Bratslavsky, E., Muraven, M., & Tice, D.M. (1998). Ego depletion:
Is the active self a limited resource? Journal of Personality and Social Psychology, 74, 1252–
1265.
Beaulieu, J.-M., & Gainetdinov, R.R. (2011). The physiology, signaling, and
pharmacology of dopamine receptors. Pharmacological Reviews, 63, 182-217.
Bialystok, E., & Craik, F.I.M. (2010). Cognitive and linguistic processing in the
bilingual mind. Current Directions in Psychological Sciences, 19, 19–23.
Bialystok, E., Craik, F.I.M., Klein, R., & Viswanathan, M. (2004). Bilingualism, aging,
and cognitive control: Evidence from the Simon task. Psychology and Aging, 19, 290–303.
Boesak, A. (1984). Apartheid, liberation and the Calvinist tradition. Johannesburg:
Skotaville Publications.
Bogacz, R. (2007). Optimal decision-making theories: linking neurobiology with
behavior. Trends in Cognitive Sciences, 11, 118–125.
Botvinick, M.M., Braver, T.S., Carter, C.S., Barch, D.M., & Cohen, J.D. (2001).
Conflict monitoring and cognitive control. Psychological Review, 108, 624–652.
Brehm, J.W., & Self, E.A. (1989). The intensity of motivation. Annual Review of
Psychology, 40, 109–131.
Chmiel, N. (1984). Phonological encoding for reading: The effect of concurrent
articulation in a Stroop task. British Journal of Psychology, 75, 213-220.
Colzato, L.S., Bajo, M.T., van den Wildenberg, W.P.M., Paolieri, D., Nieuwenhuis,
40
S.T., La Heij, W., & Hommel, B. (2008). How does bilingualism improve executive control?
A comparison of active and reactive inhibition mechanisms. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 34, 302-312.
Colzato, L.S., de Bruijn, E., & Hommel, B. (2012). Up to "me" or up to "us"? The
impact of self-construal priming on cognitive self-other integration. Frontiers in Psychology,
3:341.
Colzato, L.S., Hommel, B., & Shapiro, K. (2010). Religion and the Attentional Blink:
Depth of faith predicts depth of the blink. Frontiers in Psychology, 1:147.
Colzato, L.S., Hommel, B., van den Wildenberg, W., & Hsieh, S. (2010). Buddha as an
eye opener: A link between prosocial attitude and attentional control. Frontiers in Psychology,
1:156.
Colzato, L.S., Ozturk, A., & Hommel, B. (2012). Meditate to create: The impact of
focused-attention and open-monitoring training on convergent and divergent thinking.
Frontiers in Psychology, 3:116.
Colzato, L.S., Pratt, J., & Hommel, B. (2010). Dopaminergic control of attentional
flexibility: Inhibition of return is associated with the dopamine transporter gene (DAT1).
Frontiers in Neuroscience, 4:53.
Colzato, L.S., Slagter, H., de Rover, M., & Hommel, B. (2011). Dopamine and the
management of attentional resources: Genetic markers of striatal D2 dopamine predict
individual differences in the Attentional Blink. Journal of Cognitive Neuroscience, 23, 35763585.
Colzato, L.S., Szapora, A., Lippelt, D., & Hommel, B. (in press). Prior meditation
practice modulates performance and strategy use in convergent- and divergent-thinking
problems. Mindfulness.
Colzato, L.S., van Beest, I., van den Wildenberg, W.P.M., Scorolli, C., Dorchin, S.,
41
Meiran, N., Borghi, A.M., & Hommel, B. (2010). God: Do I have your attention? Cognition,
117, 87-94.
Colzato, L.S., van den Wildenberg, W.P.M., & Hommel, B. (2008). Losing the big
picture: How religion may control visual attention. PLoS ONE 3(11): e3679.
Colzato, L.S., van den Wildenberg, W.P.M., & Hommel, B. (2013). Increasing selfother integration through divergent thinking. Psychonomic Bulletin & Review, 20, 1011-1016.
Colzato, L.S., van den Wildenberg, W.P.M., & Hommel, B. (2014). Cognitive control
and the COMT Val158Met polymorphism: Genetic modulation of videogame training and
transfer to task-switching efficiency. Psychological Research, 78, 670-678.
Colzato, L.S., van den Wildenberg, W.P.M., van der Does, W., & Hommel, B. (2010).
Genetic markers of striatal dopamine predict individual differences in dysfunctional, but not
functional impulsivity. Neuroscience, 170, 782-788.
Colzato, L.S., van Hooidonk, L., van den Wildenberg, W.P.M., Harinck, F., & Hommel,
B. (2010). Sexual orientation biases attentional control: a possible gaydar mechanism.
Frontiers in Psychology, 1:13.
Colzato, L.S., van Leeuwen, P.J.A., van den Wildenberg, W., & Hommel, B. (2010).
DOOM’d to switch: Superior cognitive flexibility in players of first person shooter games.
Frontiers in Psychology, 1:8.
Colzato, L.S., Waszak, F., Nieuwenhuis, S., Posthuma, D., & Hommel, B. (2010). The
flexible mind is associated with the Catechol-O-methyltransferase (COMT) Val158Met
polymorphism: Evidence for a role of dopamine in the control of task switching.
Neuropsychologia, 48, 2764-2768.
Colzato, L.S., Zech, H., Hommel, B., Verdonschot, R., van den Wildenberg, W., &
Hsieh, S. (2012). Loving-kindness brings loving-kindness: The impact of Buddhism on
cognitive self-other integration. Psychonomic Bulletin & Review, 19, 541-545.
42
Colzato, L.S., Zmigrod, S., & Hommel, B. (2013). Dopamine, norepinephrine, and the
management of sensorimotor bindings: Individual differences in updating of stimulus-response
episodes are predicted by DAT1, but not DBH5'-ins/del. Experimental Brain Research, 228,
213-220.
Cools, R. (2008). Role of dopamine in the motivational and cognitive control of
behaviour. Neuroscientist 14, 381–395.
Cools, R. (2012). Chemical neuromodulation of goal-directed behavior. In: P.M. Todd,
T.T. Hills, & T.W. Robbins (eds.), Cognitive search: Evolution, algorithms, and the brain:
Strüngmann Forum Report. Cambridge, MA: MIT Press.
Cools, R., & D’Esposito (2010). Dopaminergic modulation of flexible cognitive control
in humans. In: A. Björklund, S. Dunnett, L. Iversen, & S. Iversen (eds.), Dopamine handbook
(pp. 249-260). Oxford: Oxford University Press.
Cools, R., & D’Esposito, M. (2011). Inverted U-shaped dopamine actions on human
working memory and cognitive control. Biological Psychiatry, 69:e113-25.
Desimone, R., & Duncan, J. (1995). Neural mechanism of selective visual attention.
Annual Review of Neuroscience, 18, 193-222.
de Wit, S. & Dickinson, A. (2009). Associative theories of goal-directed behaviour. A
case for animal-human translational models. Psychological Research, 73, 463–476.
Dogen, I. (1976). Shobogenzo (Y.Yokai, Trans.). NewYork: Weatherhill.
Dolk, T., Hommel, B., Colzato, L.S., Schütz-Bosbach, S., Prinz, W., & Liepelt, R.
(2014). The joint Simon effect: A review and theoretical integration. Frontiers in Psychology,
5:974.
Dreisbach, G. & Goschke, T. (2004). How positive affect modulates cognitive control:
Reduced perseveration at the cost of increased distractibility. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 30, 343-353.
43
Durstewitz, D., & Seamans, J.K. (2008). The dual-state theory of prefrontal cortex
dopamine function with relevance to catechol-o-methyltransferase genotypes and
schizophrenia. Biological Psychiatry, 64, 739-749.
Egner, T. (2007). Congruency sequence effects and cognitive control. Cognitive,
Affective, & Behavioral Neuroscience, 7, 380-390.
Egner, T., & Hirsch, J. (2005). Cognitive control mechanisms resolve conflict through
cortical amplification of task-relevant information. Nature Neuroscience, 8, 1784–1790.
Exner, S. (1879). Physiologie der Grosshirnrinde. In L. Hermann (Ed.), Handbuch der
Physiologie, 2. Band, 2. Theil (pp. 189-350). Leipzig: Vogel.
Fischer, R., & Hommel, B. (2012). Deep thinking increases task-set shielding and
reduces shifting flexibility in dual-task performance. Cognition, 123, 303-307.
Förster, J. (2012). GLOMOsys: The how and why of global and local processing.
Current Directions in Psychological Science, 21, 15-19.
Freud, S. (1923). Das Ich und das Es. Leipzig etc.: Internationaler Psycho-analytischer
Verlag. English translation: The Ego and the Id, Joan Riviere (trans.), London: Hogarth Press
and Institute of Psycho-analysis.
Friedman, N.P., Miyake, A., Young, S.E., DeFries, J.C., Corley, R.P., & Hewitt, J.K.
(2008). Individual differences in executive functions are almost entirely genetic in origin.
Journal of Experimental Psychology: General, 137, 201–225.
Goschke, T. (2003). Voluntary action and cognitive control from a cognitive
neuroscience perspective. In S. Maasen, W. Prinz, & G. Roth (Hrsg.), Voluntary action: Brains,
minds, and sociality (pp.49-85). Oxford: Oxford University Press.
Gratton, G., Coles, M.G.H., & Donchin, E. (1992). Optimizing the use of information:
Strategic control of activation of responses. Journal of Experimental Psychology. General, 121,
480–506.
44
Greenwald, A.G., Banaji, M.R., Rudman, L.A., Farnham, S.D., Nosek, B.A., & Mellott,
D.S. (2002). A unified theory of implicit attitudes, stereotypes, self-esteem, and self-concept.
Psychological Review, 109, 3–25.
Guilford, J.P. (1967). The nature of human intelligence. New York: McGraw-Hill.
Hillgruber, A. (1912). Fortlaufende Arbeit und Willensbetätigung [Continuous work
and will performance]. Untersuchungen zur Psychologie und Philosophie, 1.
Hommel, B. (2000a). Intentional control of automatic stimulus-response translation. In
Y. Rossetti & A. Revonsuo (eds.), Interaction between dissociable conscious and
nonconscious processes (pp. 223-244). Amsterdam: John Benjamins Publishing Company.
Hommel, B. (2000b). The prepared reflex: Automaticity and control in stimulusresponse translation. In S. Monsell & J. Driver (eds.), Control of cognitive processes: Attention
and performance XVIII (pp. 247-273). Cambridge, MA: MIT Press.
Hommel, B. (2009). Action control according to TEC (theory of event coding).
Psychological Research, 73, 512-526.
Hommel, B. (2012). Convergent and divergent operations in cognitive search. In: P.M.
Todd, T.T. Hills, & T.W. Robbins (eds.), Cognitive search: Evolution, algorithms, and the
brain (pp. 221-235). Cambridge, MA: MIT Press.
Hommel, B. (2013). Ideomotor action control: On the perceptual grounding of
voluntary actions and agents. In W. Prinz, M. Beisert & A. Herwig (Eds.), Action science:
Foundations of an emerging discipline (pp. 113-136). Cambridge, MA: MIT Press.
Hommel, B., & Colzato, L.S. (2010). Religion as a control guide: on the impact of
religion on cognition. Zygon: Journal of Religion & Science, 45, 596-604.
Hommel, B., Colzato, L.S., Fischer, R., & Christoffels, I. (2011). Bilingualism and
creativity: Benefits in convergent thinking come with losses in divergent thinking. Frontiers in
Psychology, 2:273.
45
Hommel, B., Colzato, L.S., Scorolli, C., Borghi, A.M., & van den Wildenberg, W.P.M.
(2011). Religion and action control: Faith-specific modulation of the Simon effect but not stopsignal performance. Cognition, 120, 177-185.
Hommel, B., Colzato, L.S., & van den Wildenberg, W.P.M. (2009). How social are task
representations? Psychological Science, 20, 794-798.
Hommel, B., Fischer, R., Colzato, L.S., van den Wildenberg, W.P.M., & Cellini, C.
(2012). The effect of fMRI (noise) on cognitive control. Journal of Experimental Psychology:
Human Perception and Performance, 38, 290-301.
Hommel, B., Kessler, K., Schmitz, F., Gross, J., Akyürek, E., Shapiro, K., & Schnitzler,
A. (2006). How the brain blinks: Towards a neurocognitive model of the Attentional Blink.
Psychological Research, 70, 425-435.
Hommel, B., Müsseler, J., Aschersleben, G., & Prinz, W. (2001). The theory of event
coding (TEC): A framework for perception and action planning. Behavioral and Brain
Sciences, 24, 849-878.
Isen, A.M. (1999). On the relationship between affect and creative problem solving. In
S. Russ (Ed.), Affect, creative experience, and psychological adjustment (pp. 3–17). London:
Taylor & Francis.
Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall.
Kiesel, A., Steinhauser, M., Wendt, M., Falkenstein, M., Jost, K., Philipp, A.M., &
Koch, I. (2010). Control and interference in task switching - A review. Psychological Bulletin,
136, 849-874.
Kim, D., & Hommel, B. (in press). An event-based account of conformity.
Psychological Science.
Kornblum, S., Hasbroucq, T., & Osman, A. (1990). Dimensional overlap: Cognitive
basis for stimulus-response compatibility - a model and taxonomy. Psychological Review, 97,
46
253-270.
Kukla, A. (1972). Foundations of an attributional theory of performance. Psychological
Review, 79, 454–470.
La Heij...
Lippelt, D.P., Hommel, B., & Colzato, L.S. (2014). Focused attention, open monitoring
and loving kindness meditation: Effects on attention, conflict monitoring and creativity.
Frontiers in Psychology, 5:1083.
Logan, G.D., & Cowan, W.B. (1984). On the ability to inhibit thought and action: A
theory of an act of control. Psychological Review, 91, 295–327.
Logan, G.D., & Gordon, R.D. (2001). Executive control of visual attention in dual-task
situations. Psychological Review, 108, 393-434.
Lutz, A., Slagter, H.A., Dunne, J.D., & Davidson, R.J. (2008). Attention regulation and
monitoring in meditation. Trends in Cognitive Sciences, 12, 163–169.
MacLeod, C.M. (1991). Half a century of research on the Stroop effect: An integrative
review. Psychological Review, 109, 163-203.
Martens, S., Munneke, J., Smid, H., & Johnson, A. (2006). Quick minds don't blink:
Electrophysiological correlates of individual differences in attentional selection. Journal of
Cognitive Neuroscience, 18, 1423-1438.
Martin, M. (1978). Speech recoding in silent reading. Memory & Cognition, 30, 187200.
Mayr, U., & Keele, S. W. (2000). Changing internal constraints on action: The role of
backward inhibition. Journal of Experimental Psychology: General, 129, 4–26.
McCauley, R.N. (1987). The role of cognitive explanations in psychology.
Behaviorism, 15, 27-40.
McClain, L. (1983). Effects of response type and set size on Stroop color-word
47
performance. Perceptual & Motor Skills, 56, 735-743.
Mednick, S.A. (1968). The Remote Associates Test. The Journal of Creative Behavior,
2, 213–214.
Michotte, A., & Prüm, E. (1911). Etude experimental sur la choix volontaire. Archives
de Psychologie, 10, 113-320.
Miyake, A., Friedman, N.P., Emerson, M.J., Witzki, A.H., Howerter, A., Wager, T.
(2000). The unity and diversity of executive functions and their contributions to complex
"frontal lobe" tasks: A latent variable analysis. Cognitive Psychology, 41, 49–100.
Moors, A., & De Houwer, J. (2006). Automaticity: A theoretical and conceptual
analysis. Psychological Bulletin, 132, 297-326.
Müller, J., Dreisbach, G., Goschke, T., Hensch, T., Lesch, K.-P. & Brocke, B. (2008).
Dopamine and cognitive control: The prospect of monetary gains influences the balance
between flexibility and stability in a set-shifting paradigm. European Journal of Neuroscience,
26, 3661–3668.
Navon, D. (1977). Forest before trees: The precedence of global features in visual
perception. Cognitive Psychology, 9, 353–383.
Navon, D. (1984). Resources—a theoretical soup stone? Psychological Review, 91,
216-234.
Nijstad, B.A., De Dreu, C.K.W., Rietzschel, E.F., & Baas, M. (2010). The dual pathway
to creativity model: Creative ideation as a function of flexibility and persistence. European
Review of Social Psychology, 21, 34–77.
Nisbett, R.E., & Miyamoto, Y. (2005). The influence of culture: holistic versus analytic
perception. Trends in Cognitive Sciences, 9, 467–473.
Nolan, K., Bilder, R., Lachman, H., & Volavka, K. (2004). Catechol-Omethyltransferase Val158Met polymorphism in schizophrenia: Differential effects of Val and
48
Met alleles on cognitive stability and flexibility. American Journal of Psychiatry, 161, 359–
361.
Olivers, C.N.L., & Nieuwenhuis, S. (2006). The beneficial effects of additional task
load, positive affect, and instruction on the attentional blink. Journal of Experimental
Psychology: Human Perception and Performance, 32, 364-379.
Ratcliff, R., & Rouder, J.N. (1998). Modeling response times for two-choice decisions.
Psychological Science, 9, 347-356.
Reuss, H., Kiesel, A., Kunde, W., & Hommel, B. (2011). Unconscious activation of
task sets. Consciousness and Cognition, 20, 556-567.
Reuter, D.F. (2002). Gaydar: The ultimate insider guide to the gay sixth sense. New
York: Crown.
Ridderinkhof, K.R., Forstmann, B.U., Wylie, S., Burle, B., & van den Wildenberg,
W.P.M. (2010). Neurocognitive mechanisms of action control: Resisting the call of Sirens.
Wylie Interdisciplinary Reviews (WIREs) Cognitive Science, 2, 174-192.
Ryle, G. (1949). The concept of mind. Chicago: The University of Chicago Press.
Sellaro, R., Hommel, B., de Kwaadsteniet, E.W., & Colzato, L.S. (2014). Increasing
interpersonal trust through divergent thinking. Frontiers in Psychology, 5:561.
Sebanz, N., Knoblich, G., & Prinz, W. (2003). Representing others’ actions: Just like
one’s own? Cognition, 88, B11–B21.
Sellaro, R., van Dijk, W.W., Rossi Paccani, C., Hommel, B., Colzato, L.S., (2014). A
question of scent: lavender aroma promotes interpersonal trust. Frontiers in Psychology,
5:1486.
Shapiro, K., Schmitz, F., Martens, S., Hommel, B., & Schnitzler, A. (2006). Resource
sharing in the attentional blink. Neuroreport, 17, 163-166.
Stock, A., & Stock, C. (2004). A short history of ideo-motor action. Psychological
49
Research, 68, 176–188.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of
Experimental Psychology, 28, 643–662.
van Steenbergen, H., Band, G.P.H., & Hommel, B. (2009). Reward counteracts conflict
adaptation: Evidence for a role of affect in executive control. Psychological Science, 20, 14731477.
van Steenbergen, H., Band, G.P.H., & Hommel, B. (2010). In the mood for adaptation:
How affect regulates conflict-driven control. Psychological Science, 21, 1629-1634.
van Steenbergen, H., Band, G.P.H., & Hommel, B. (2012). Reward valence modulates
conflict-driven attentional adaptation: Electrophysiological evidence. Biological Psychology,
90, 234-241.
van Steenbergen, H., Band, G.P.H., Hommel, B., Rombouts, S.A.R.B., & Nieuwenhuis,
S. (in press). Hedonic hotspots regulate cingulate-driven adaptation to cognitive demands.
Cerebral Cortex.
van Steenbergen, H., Booij, L., Band, G.P.H., Hommel, B., & van der Does, A.J.W.
(2012). Affective regulation of conflict-driven control in remitted depressive patients after
acute tryptophan depletion. Cognitive, Affective, & Behavioral Neuroscience, 12, 280-286.
van Steenbergen, H., Langeslag, S.J.E., Band, G.P.H., & Hommel, B. (2014). Reduced
cognitive control in passionate lovers. Motivation and Emotion, 38, 444-450.
van Veen, V., & Carter, C.S. (2002). Conflict and cognitive control in the brain. Current
Directions in Psychological Science, 15, 237-240.
Verbruggen, F., McLaren, I.P.L., & Chambers, C.D. (2014). Banishing the control
homunculi in studies of action control and behavior change. Perspectives on Psychological
Science, 9, 497-524.
Wallas, G. (1926). The art of thought. New York: Harcourt Brace.
50
Waszak, F., Hommel, B., & Allport, A. (2003). Task-switching and long-term priming:
Role of episodic stimulus-task bindings in task-shift costs. Cognitive Psychology, 46, 361-413.
Wickens, C.D. (1984). Processing resources in attention. In: R. Parasuraman & R.
Davies (eds.), Varieties of attention (pp. 63-10l). New York: Academic Press.
Wiers, R.W., & Stacy, A.W. (2006) (eds), Handbook of implicit cognition and
addiction. Thousand Oaks, CA: Sage.
Zwosta, K., Hommel, B., Goschke, T., & Fischer, R. (2013). Mood states determine the
degree of task shielding in dual-task performance. Cognition & Emotion, 27, 1142-1152.
11. Acknowledgments
This research was supported by a grant of the Netherlands Research Organization
(NWO) to the author (433-09-243).
51
Figure captions
Figure 1. Possible mechanisms involved in competitive decision-making. The goalrelated Alternative 1 is supported by the goal representation (which provides a “top-down
bias”), but competes with choice Alternative 2 for selection through mutual inhibition
(“competition”). Selection requires one alternative to pass a certain threshold (not shown),
which given the mutual inhibition would result in the relative inhibition of the non-selected
alternative. Figure 1A shows the generic model; Figures 1B and 1D show how strong
persistence and pronounced flexibility might be achieved by modulating the strength of the
top-down support from goal representations, respectively; Figures 1C and 1E show how strong
persistence and pronounced flexibility might be achieved by modulating the strength of mutual
inhibition between alternatives, respectively; and Figures 1F and 1G shows the assumptions
underlying the discussion in this chapter, which combine both possible mechanisms (goal
modulation and the modulation of mutual inhibition).
Alternative 1
(topdown
bias)
Goal
Alternative 2
(competition)
A
Alternative 1
Goal
Alternative 2
Alternative 1
Goal
Alternative 2
Flexibility
Alternative 2
Persistence
Alternative 1
E
Alternative 2
C
Alternative 1
Goal
Flexibility
Persistence
Goal
D
B
Alternative 1
Alternative 2
Alternative 1
Goal
Alternative 2
Flexibility
Persistence
Goal
G
F